Key Points
- Anthropic received an official “supply chain risk” classification from the Pentagon with immediate effect.
- Government contractors face prohibition on deploying Claude AI for Defense Department projects.
- Military operations in Iran and Venezuela reportedly utilized Claude technology.
- CEO Dario Amodei announced plans to pursue legal action against the classification.
- This designation usually targets foreign adversaries, such as Huawei from China.
The Department of Defense has assigned Anthropic a supply chain risk classification, effectively preventing government contractors from deploying the company’s Claude AI system in Pentagon projects.
This immediate designation places Anthropic alongside entities typically considered national security threats. China’s Huawei stands as the most prominent example of a company receiving this classification.
Anthropic CEO Dario Amodei addressed the designation through a public announcement. He clarified the restriction’s limited scope, explaining it exclusively affects direct Pentagon contractual work rather than all Claude usage by companies holding military contracts.
“It plainly applies only to the use of Claude by customers as a direct part of contracts with the Department of War,” Amodei wrote.
Claude has achieved significant integration within U.S. military infrastructure. Sources with knowledge of the situation report Claude’s deployment in military operations targeting Iran and Venezuela, where it supported intelligence analysis and operational planning activities.
Extracting the technology presents considerable challenges. Industry analysts characterize the removal process as “painful” due to Claude’s extensive integration across military systems.
Origins of the Conflict
Tension between Anthropic and Pentagon leadership has intensified over several months. The core issue centers on disagreements regarding safety protocols.
Anthropic maintains restrictions preventing Claude from enabling autonomous weapons systems or large-scale domestic surveillance programs. Pentagon officials contended they should possess unrestricted access to the technology, provided usage remains within legal boundaries.
Public awareness of the dispute emerged earlier this year before reaching critical levels this week. The Information published an internal Anthropic memo Wednesday that Amodei originally wrote the previous Friday, intensifying tensions. The memo included Amodei’s suggestion that Pentagon officials harbored negative feelings toward Anthropic partly because “we haven’t given dictator-style praise to Trump.”
Amodei issued an apology regarding the memo’s publication. Reports indicate Anthropic investors actively worked to mitigate fallout from the controversy.
Pentagon Chief Technology Officer Emil Michael used X Thursday evening to announce the absence of active Defense Department negotiations with Anthropic.
Future Implications
Amodei revealed in his statement that Anthropic and Pentagon representatives had explored potential arrangements allowing military collaboration while maintaining the company’s safety protocols. These discussions have yet to produce an agreement.
Amodei confirmed Anthropic’s intention to pursue legal challenges against the designation.
Microsoft conducted its own assessment of the designation, determining Claude remains accessible to customers through platforms including Microsoft 365, GitHub, and its AI Foundry—with the exception of Department of War contracts.
Palantir’s Maven Smart Systems, which delivers intelligence analysis and weapons targeting capabilities to military organizations, had developed multiple operational workflows incorporating Anthropic’s Claude code.
Amazon, holding significant investment stakes in Anthropic, had provided no statement by publication time.
